Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 1 Niche filtering, not interspecific resource competition, explains the co-occurrence 2 of butterfly species across the Japanese archipelago 3 4 5 6 7 8 9 10 11 12 13 Ryosuke Nakadai12*†, Koya Hashimoto1*, Takaya Iwasaki3, Yasuhiro Sato14 1 Center for Ecological Research, Kyoto University, Hirano 2-509-3, Otsu, Shiga, 520-2113 Japan 2 Current address: Faculty of Science, University of the Ryukyus, Senbaru 1, Nishihara, Okinawa, 903-0213 Japan. 3 Department of Biological Sciences, Faculty of Science, Kanagawa University, Tsuchiya 2946, Hiratsuka, Kanagawa 259-1293 Japan 4 Current address: Department of Plant Life Sciences, Faculty of Agriculture, Ryukoku University, Yokotani 1-5, Otsu, Shiga 520-2194 Japan. 14 15 * Equal contribution 16 17 † Author Correspondence: R. Nakadai 18 Faculty of Science, University of the Ryukyus, Senbaru 1, Nishihara, Okinawa 903-0213, 19 Japan. 20 E-mail: [email protected] 21 22 Running head: Co-occurrence of butterfly species 23 24 Keywords: butterfly, co-occurrence, dispersal ability, climatic niche, herbivorous insect, 25 host use, taxonomic relatedness 26 27 bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 28 Abstract 29 The relevance of interspecific resource competition in the context of community 30 assembly by herbivorous insects is a well-known topic in ecology. Most previous 31 studies focused on local species assemblies, that shared host plants. Few studies 32 evaluated species pairs within a single taxon when investigating the effects of host plant 33 sharing at the regional scale. Herein, we explore the effect of plant sharing on the 34 geographical co-occurrence patterns of 229 butterflies distributed across the Japanese 35 archipelago; we use two spatial scales (10 × 10 km and 1 × 1 km grids) to this end. We 36 considered that we might encounter one of two predictable patterns in terms of the 37 relationship between co-occurrence and host-sharing among butterflies. On the one 38 hand, host-sharing might promote distributional exclusivity attributable to interspecific 39 resource competition. On the other hand, sharing of host plants might promote 40 co-occurrence attributable to filtering by resource niche. At both grid scales, we found 41 significant negative correlations between host use similarity and distributional 42 exclusivity. Our results support the thesis that the butterfly co-occurrence pattern across 43 the Japanese archipelago is better explained by filtering via resource niche rather than 44 interspecific resource competition. 45 46 Introduction 47 Efforts to understand community assembly processes are of major importance in 48 ecological research (Diamond 1975, Cavender-Bares et al. 2009). All of dispersal 49 limitations, environmental filtering via both abiotic and biotic niches, and interspecific 50 interactions are thought sequentially to determine local community structures (reviewed 51 by Cavender-Bares et al. 2009). Of the various relevant factors, the significance of 52 interspecific interaction has often been assessed by examining how different species 53 co-occur spatially even when they share similar niches (Diamond 1975, Gotelli and 54 McCabe 2002). As different species with similar niches likely prefer similar bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 55 environmental habitats, but may compete strongly, recent studies have often compared 56 the significance of interspecific interactions (in terms of community assembly patterns) 57 from the viewpoint of niche filtering (Webb et al. 2002, Mayfield and Levine 2010). 58 In terrestrial ecosystems, many plant species are used by various herbivorous 59 insects as both food resources and habitats, suggesting that host use plays key roles in 60 both competition and filtering via resource niches. Earlier studies described niche 61 partitioning between co-occurring herbivores and considered that partitioning was 62 attributable to interspecific competition (Ueckert and Hansen 1971, Benson 1978, 63 Waloff 1979, Augustyn et al. 2016). Conversely, other studies described the frequent 64 co-occurrence of multiple herbivorous insect species on shared host plants despite the 65 fact that extensive niche overlap was in play (Ross 1957, Rathcke 1976, Bultman and 66 Faeth 1985, Hashimoto and Ohgushi 2017); the cited authors argued that the effects of 67 interspecific competition were relatively weak in terms of community organization 68 (Lawton and Strong 1981, Schoener 1983, Strong et al. 1984, Tack et al. 2009). Whether 69 the importance of competition was or was not supported, most previous studies focused 70 on how host plants shape the local assemblages of herbivorous insects. Therefore, the 71 question of whether interspecific resource competition is critical in terms of shaping 72 herbivore communities remains highly controversial (reviewed by Kaplan and Denno 73 2007). Very little is known about the extent to which the results of the cited studies can 74 be extrapolated to describe patterns, at the regional scale, of the distribution of 75 herbivorous insects within a single taxon. 76 The effects of interspecific competition and filtering via resource niches on 77 the co-occurrence patterns of herbivorous insects must be considered when attempting 78 to explain differences in species traits, with the exception of host use. For example, 79 climatic niches (often associated with differences in potential geographical distributions 80 in the absence of interspecific interactions; Warren et al. 2008, Takami and Osawa 81 2016) may drive niche filtering, which may in turn mediate the impact of host use on bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 82 assembly patterns. In addition, taxonomic relatedness should reflect niche similarity; the 83 niche of any organism should be partly determined by its phylogenetic history (Webb et 84 al. 2002). Thus, the outcomes of competition within shared niches should reflect both 85 resource and climatic factors (Mayfield and Levine 2010). As such factors may 86 correlate with host use by individual species, any focus on host use alone may yield 87 misleading results. Furthermore, dispersal ability may complicate assembly patterns; 88 these are often associated with both the extent of the geographical range and other 89 factors (Kneitel and Chase 2004). A high dispersal ability may enhance the extent of 90 co-occurrence. Thus, when attempting to explore the effects of competition and filtering 91 via resource niches on distribution patterns, it is important to consider all of taxonomic 92 relatedness, climatic niche preferences, and dispersal ability. 93 In the present study, we explored whether interspecific competition or niche 94 filtering better explained the geographical co-occurrence of a group of herbivorous 95 insects. We expected to discern one of two patterns when evaluating the significance of 96 interspecific competition in terms of the geographical patterns of species co-occurrence. 97 One possibility was that sharing of host plants would be associated with exclusive 98 distributions attributable to resource competition. The alternative was that sharing of 99 host plants would promote species co-occurrence attributable to filtering via resource 100 niches. The former pattern predicts that a positive correlation would be evident between 101 host use similarity (i.e., the extent of sharing of host species) and the exclusiveness of 102 geographic distribution. The latter pattern predicts that the correlation would be 103 negative. We focused on herbivorous butterflies of the Japanese archipelago, for the 104 following reasons. First, butterfly host specificity is relatively high; the insect larvae are 105 mostly leaf chewers (Novotny et al. 2010). Second, the Biodiversity Center of Japan has 106 extensive records of butterfly distributions. Third, forewing length (easily measured on 107 photographs) is a useful index of butterfly dispersal ability (Chai and Srygley 1990; 108 Shirôzu 2006). Finally, and most importantly, a great deal is known about the hosts bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 109 used by butterfly species (Saito et al. 2016). Thus, data on Japanese butterflies can be 110 used to explore the effects of host plant sharing and other factors on the co-occurrence 111 patterns at regional scales. Specifically, we determined the effect of host use similarity 112 on the exclusivity of geographical distribution between pairs of entirely herbivorous 113 Japanese butterflies; we considered taxonomic relatedness, climatic niche similarities, 114 and overall dispersal abilities in the course of our work. 115 116 Methods 117 Study area 118 The Japanese archipelago, including the Ryukyu Islands, forms a long chain of 119 continental islands lying off the eastern coast of Asia. The latitudinal range of the 120 archipelago (22°N to 45°N) embraces hemi-boreal, temperate, and subtropical zones. 121 The mean temperatures in the coldest and warmest months are −19.0°C and 31.5°C, 122 respectively; the annual precipitation ranges from 867 to 3,908 mm (Kubota et al. 123 2014). Study organisms 124 125 The Japanese archipelago hosts over 280 species of butterflies of five families (the 126 Papilionidae, Pieridae, Lycaenidae, Nymphanidae, and Hesperiidae) (Shirôzu 2006). 127 Over 95% of the larvae of Japanese butterflies feed on plants (Honda 2005). The host 128 plants are diverse and include both dicots and monocots (Saito et al. 2016). Lycaenid 129 butterflies include two non-herbivorous species, but all species of all other families are 130 exclusively herbivorous. 131 Metadata compilation 132 Butterfly census data are available on the website of the Biodiversity Center of Japan, 133 Ministry of the Environment (http://www.biodic.go.jp/index_e.html). We used the 134 results of the fourth and fifth censuses (1988 to 1991 and 1997 to 1998, respectively) of 135 the National Survey of the Natural Environment in Japan bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 136 (http://www.biodic.go.jp/kiso/15/do_kiso4.html). This database includes records of 273 137 species/subspecies of butterflies from the entire Japanese archipelago, in grid cells of 138 latitude 5 min and longitude 7.5 min (the Japanese Standard Second Mesh). These grid 139 dimensions are about 10 km × 10 km, and this grid is described below as the “10-km 140 grid.” Furthermore, the Biodiversity Center also contains records from grid cells of 141 latitude 30 s and longitude 45 s (the Japanese Standard Third Mesh). These grid 142 dimensions are about 1 km × 1 km, and this grid is described below as the “1-km grid.” 143 As processes driving community assemblies may vary between spatial scales 144 (Cavender-Bares et al. 2009), we evaluated data from both grids. We summarized data 145 at the species level, and converted all records into the presence or absence (1/0) of a 146 species in each grid. We used the taxonomy of Shirôzu (2006). 147 Data on host plants and forewing length were evaluated as possible variables 148 explaining, respectively, host use and dispersal ability. The host plants of 278 butterfly 149 species/subspecies were obtained from the data of Saito et al. (2016). Dispersal ability 150 was evaluated by reference to adult wing length. We compiled wing data on 284 species 151 using published illustrations (Shirôzu 2006). We used Image J software (Abramoff et al. 152 2004) to extract forewing lengths (in cm) from plates that included centimeter scale bars. 153 Multiple forewing lengths were extracted when individual, sexual, and/or geographical 154 variations were evident. 155 We assembled data on the distributions, host plants, climatic niches (described 156 below), and forewing lengths of 229 butterfly species. Twenty-four species were 157 excluded from the analysis, for the following reasons. First, the taxonomic status of 158 three species (Papilio dehaanii, Pieris dulcinea, and Eurema hecabe) changed in the 159 interval between the fourth and fifth biodiversity censuses (Inomata 1990; Shirôzu 160 2006). Thus, we excluded these species because identifications were unreliable. Second, 161 we excluded three species of non-herbivorous butterflies (Taraka hamada, Spindasis 162 takanonis, and Niphanda fusca). Finally, we excluded a further 18 species because the bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 163 models used to evaluate their ecological niches failed to satisfy the criteria that we 164 imposed (the details are in Appendix S2). 165 Data analysis 166 Species distribution exclusiveness. We used the checkerboard scores (C-scores; Stone 167 and Roberts 1990) to evaluate the exclusivity of distributions between each species pair. 168 We set ri and rj as the numbers of grids in which species i and j, respectively, were 169 present; the checker unit Cij associated with the two species was defined as: Cij = (ri − 170 Sij) × (rj − Sij), where Sij indicates the extent of co-presence (i.e., the number of grid 171 cells shared by the two species). Thus, the checker unit became larger as the two species 172 occurred more commonly in different grid cells. We simulated null models to allow the 173 observed checker units to be compared with stochastic distributions. We used the 174 method of Jonsson et al. (2001) to describe the frequencies of species occurrence and 175 randomized the presence/absence matrices for each pair of butterfly species. The null 176 models were run 999 times for each species pair. Cobs and Cnull were the checker units of 177 the observed and null distributions, respectively; the checker unit was standardized as: 178 Cstd = (Cobs − Cnull)/SDnull, where SDnull indicates the standard deviation of all checker 179 units of the null models. The checker unit of the null model, Cnull, was the average 180 checker unit of all null models. Thus, positive and negative values of Cstd indicate that 181 two species are allopatrically and sympatrically distributed, respectively, to extents 182 greater than indicated by the null models. All statistical analyses were performed with 183 the aid of R software version 3.2.0 (R Core Team 2015). 184 Climatic niche similarities. We used ecological niche modeling (ENM) 185 (Franklin 2010) to evaluate climatic niche similarities among butterfly species (Warren 186 et al. 2008). ENM associates distributional data with environmental characteristics, thus 187 estimating the response functions of, and the contributions made by, environmental 188 variables. Furthermore, potential distributional ranges may be obtained by projecting 189 model data onto geographical space. In the present study, potential distributional ranges bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 190 estimated by ENM should be influenced by abiotic environmental variables alone 191 (climate and altitude); we did not consider interspecific interactions among butterflies or 192 dispersal abilities in this context. Thus, comparisons of potential distribution patterns 193 estimated by ENM allow evaluation of climatic niche similarities among butterfly 194 species. The maximum entropy algorithm implemented in MaxEnt ver. 3.3.3e software 195 (Phillips et al. 2006) was employed in ENM analyses (Appendix S2 contains the details). 196 The logistic outputs of MaxEnt analyses, which can be regarded as presence 197 probabilities, were converted into presence/absence data with cut-offs defined by the 198 situations in which training sensitivity equaled specificity (Cantor et al. 1999). Such 199 cut-offs are known to be the most effective means of threshold selection (Liu et al. 200 2005). Finally, we then used Schoener’s (1968) D statistic to calculate climatic niche 201 similarities between pairs of butterfly species, based on the MaxEnt outputs. Px,i and Py,i 202 were the probabilities (assigned by MaxEnt) that species x and y would mesh to the 203 extent of i on a geographic scale; the climatic niche similarity between the two species 204 was defined as: Denv = 1 – 0.5 × Σ|Px,i – Py,i|, where Denv ranged from 0 (no niche 205 overlap) to 1 (completely identical niches). The probability assigned to the presence of 206 species x in grid i was Px,i = px,i / Σpx,i, where px,i was the logistic Maxent output for 207 species x in grid i. 208 Explanatory variables. We evaluated both host use similarity and other factors 209 that might explain exclusive species distributions (i.e., taxonomic relatedness). We 210 calculated the total dispersal abilities of species pairs and climatic niche similarities (as 211 explained above). Host use similarity was calculated as 1 minus Jaccard’s dissimilarity 212 index (Koleff et al. 2003) when host plant species were shared by two butterflies. The 213 taxonomic relatedness of each species pair was classified as: 2: in the same genus; 1: in 214 the same family; and 0: in different families. The total dispersal ability was calculated 215 as the sum of the ln(forewing lengths) of each species pair. 216 Statistical tests. We used the Mantel test, in which the response matrix bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 217 yielded pairwise Cstd data, and which included explanatory matrices, to examine the 218 effects of host use similarity and other factors (i.e., taxonomic relatedness, climatic 219 niche similarity, and total dispersal ability) on the exclusivity of butterfly distribution. 220 We calculated Spearman’s correlations because one of the explanatory variables 221 (taxonomic relatedness) was a rank variable. P-values were determined by running 222 9,999 permutations. In addition, when analyzing correlations between host use 223 similarity and other explanatory matrices, we ran Mantel tests with 9,999 permutations, 224 and calculated Spearman’s correlations. We used partial Mantel tests, in which the 225 response matrix yielded pairwise Cstd data, and in which a matrix of host use similarity 226 including the other three explanatory matrices served as a co-variable, to evaluate the 227 effects of confounding factors associated with host use similarity on the exclusivity of 228 butterfly distribution. We ran 9,999 permutations and calculated Spearman’s 229 correlations. We employed the vegdist function of the vegan package (Oksanen et al. 230 2015) and the mantel function of the ecodist package (Goslee and Urban 2007) 231 implemented in R. We performed all analyses using data from both the 10-km and the 232 1-km grids. 233 Results 234 The standardized checkerboard scores (Cstd values) of most species pairs were negative, 235 indicating that, in general, Japanese butterflies were more likely to co-occur than 236 expected by chance (Fig. 1). The Mantel test showed that host use similarity was 237 significantly, and negatively, correlated with the Cstd values at both geographical scales 238 (Fig. 1, Table 1a); all three explanatory variables exhibited significant negative 239 correlations with the Cstd values at both scales (Table 1a). Host use similarity was 240 significantly and positively correlated with both taxonomic relatedness and climatic 241 niche similarity, but we found no significant correlation with total dispersal ability 242 (Table 1b). The partial Mantel tests revealed negative correlations between the Cstd 243 values and host use similarity, attributable to both taxonomic relatedness and dispersal bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 244 ability, at both geographical scales (Table 1c). In contrast, we found a significant 245 positive correlation in terms of climatic niche similarity in the 10-km grid dataset, but 246 no significant correlation in the 1-km grid dataset (Table 1c). 247 248 Discussion 249 Significant negative correlations were clearly evident between the Cstd scores and host 250 use similarities at both grid scales (Fig. 1, Table 1a), indicating that pairs of Japanese 251 butterflies were more likely to share host plants than not. Significant negative 252 correlations between Cstd scores and host use similarities were evident after controlling 253 for other potentially confounding factors, with the exception of climatic niche similarity 254 (Table 1c). Interspecific resource competition (i.e., a positive correlation between the 255 Cstd score and host use similarity) was detected only after controlling for the effects of 256 climatic niche similarity in the 10-km grid data, but the correlation coefficient was low 257 (Table 1c). Our results are consistent with the idea that interspecific resource 258 competition is too weak to organize communities of herbivorous insects effectively 259 (Lawton and Strong 1981, Strong et al. 1984). Rather, our results suggest that the 260 geographic pattern of species co-occurrence among Japanese butterflies is better 261 explained by niche filtering. However, the niche axes that drive the observed patterns 262 remain unclear because the factors tested exhibited mutual correlations (Table 1b). 263 The most likely explanation of our data is that the relative strength of 264 structuring via resource competition may be weaker than that associated with niche 265 filtering. As the geographical distributions of host plants would be expected to be 266 strongly associated with the local climatic environment, the impacts of resource and 267 climatic niche filtering may combine to ensure that butterfly species sharing host plants 268 assemble in the same places. In addition, the dispersal of adult butterflies from the 269 patches in which they were born may counteract the structuring force imposed by 270 interspecific competition. Indeed, co-occurrence was facilitated by the overall total bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 271 dispersal ability (Table 1a). However, negative correlations between the Cstd scores and 272 host use similarity were evident even when we controlled for the effects of total 273 dispersal ability (Table 1c). This means that dispersal alone may not explain the weak 274 impact of resource competition on the co-occurrence patterns of Japanese butterflies. 275 Other potential factors reducing the effects of interspecific resource competition may 276 also be in play; we did not address these topics in the present study. For example, the 277 presence of natural enemies is known to reduce interspecific competition during 278 community assembly markedly (Strong 1982, Nakadai and Kawakita 2017). It is very 279 difficult to assess the effects of natural enemies at regional scales. 280 The negative correlation evident between taxonomic relatedness and the Cstd 281 scores (Table 1a) suggests that niche filtering is in play among Japanese butterflies, 282 given that taxonomic relatedness serves as a proxy of niche similarity including host use. 283 Indeed, we found significant (positive) correlations between host use similarity and the 284 taxonomic relatedness of Japanese butterflies (Table 1b), as has often been shown for 285 other herbivorous insects (e.g., Nyman et al. 2010). Moreover, when host use similarity 286 was controlled using the partial Mantel test, taxonomic relatedness did not significantly 287 affect co-occurrence at the 10-km grid scale (Appendix S1: Table S4). These results 288 suggest that, at least at the 10-km grid scale, the effects of taxonomic relatedness largely 289 reflect host use similarity. 290 In the present study, we used ENM to evaluate the effects of climatic niche 291 similarity on co-occurrence patterns. When we controlled for the effects of such niche 292 similarity, the negative correlations between the Cstd scores and host plant similarities 293 disappeared at both spatial scales (Table 1c). This suggests that the explanatory power 294 of climatic niche filtering is stronger than that of resource niche filtering. It should be 295 noted that, although ENM has been widely used to quantify climatic niches (e.g., Kozak 296 et al. 2008, Warren et al. 2008, Takami and Osawa 2016), ENM data should be treated 297 with caution (Peterson et al. 2011, Warren 2012, Warren et al. 2014). For example, bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 298 Warren et al. (2014) noted that ENM always includes non-targeted factors that limit real 299 distributions if those distributions correlate spatially with environmental predictors. 300 Such confounding effects may cause overestimation of any positive correlation between 301 climatic niche similarity and butterfly co-occurrence, and may also cause the effects of 302 host use similarity to be underestimated when controlling for the effects of climatic 303 niche similarity. The use of ENM to study species co-occurrence patterns requires 304 careful consideration; more case studies evaluating the relative importance of climatic 305 niches are required. 306 307 Conclusions 308 The significance of interspecific resource competition in terms of the structuring of 309 herbivorous insect communities is a source of long-standing controversy. Many 310 researchers have sought to explain the general patterns of relationships between the 311 co-occurrence of, and the use of different niches by, herbivorous insects. However, the 312 data remain limited because most previous studies employed narrow taxonomic and 313 spatial scales. In this context, our study is the first to provide a comprehensive picture of 314 the co-occurrence patterns among a single taxonomic group over a large region. 315 Co-occurrence of Japanese butterflies is more likely to be driven by niche filtering than 316 interspecific resource competition. It is essential to employ broad taxonomic and spatial 317 scales when attempting to reveal general patterns of community assembly among 318 herbivorous insects. Future studies should explore the relative importance of each 319 assembly stage not only ecologically but also over evolutionary time (Rabosky 2009). 320 Such work would answer the important question: “Why have herbivorous insects 321 become one of the most diverse groups of the natural world?” 322 323 Acknowledgement bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 324 We thank K. Kadowaki for his comments and advice on our manuscript. We also thank 325 the Biodiversity Center of Japan, Ministry of the Environment, Japan, for allowing 326 access to butterfly data at the 1-km grid scale. This work was supported by a grant from 327 the Grant-in-Aid Program for JSPS Fellows (grant no. 15J00601). 328 Literature cited 329 330 331 332 Abramoff, M. D., P. J. Magalhães, and S. J. Ram. 2004. Image processing with ImageJ. Biophotonics international 11:36–42. Augustyn, W. J., B. Anderson, and A. G. Ellis. 2016. Experimental evidence for 333 fundamental, and not realised, niche partitioning in a plant-herbivore community 334 interaction network. Journal of Animal Ecology 85:994–1003. 335 336 Benson, W. W. 1978. Resource Partitioning in Passion Vine Butterflies. Evolution 32:493–518. 337 Bultman, T. L. and S. H. Faeth. 1985. Patterns of intra- and inter- specific association in 338 leaf-mining insects on three oak host species. Ecological Entomology 10:121–129. 339 Cantor, S. B., C. C. Sun, G. Tortolero-Luna, R. Richards-Kortum, and M. Follen. 1999. 340 A Comparison of C/B Ratios from studies using receiver operating characteristic 341 curve analysis. Journal of Clinical Epidemiology 52:885–892. 342 Cavender-Bares, J., K. H. Kozak, P. V. A. Fine, and S. W. Kembel. 2009. The merging 343 of community ecology and phylogenetic biology. Ecology Letters 12:693–715. 344 Chai, P., and R. B. Srygley. 1990. Predation and the flight, morphology, and temperature 345 346 of neotropical rainforest butterflies. The American Naturalist 135:748–765. Diamond, J. 1975. Assembly of species communities. In: Cody ML, Diamond JM (eds) 347 Ecology and evolution of communities. Harvard University Press, Cambridge, pp 348 342–444 349 350 Franklin, J. 2010. Mapping species distributions: spatial inference and prediction. Cambridge University Press. bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 351 352 353 354 355 Goslee, S. C. and D. L. Urban. 2007. The ecodist package for dissimilarity-based analysis of ecological data. Journal of Statistical Software 22:1–19. Gotelli, N. J., and D. J. McCabe. 2002. Species co-occurrence: A meta-analysis of J. M. Diamond’s assembly rules model. Ecology 83:2091–2096. Hashimoto, K., and T. Ohgushi. 2017. How do two specialist butterflies determine 356 growth and biomass of a shared host plant? Population Ecology 59:17–27. 357 Honda, K. 2005. Larval feeding habit and host selection. In: Honda K, Kato Y (eds) 358 Biology of butterflies. University of Tokyo Press, Tokyo, pp 255–301 (in 359 Japanese). 360 Inomata, T. 1990. Keys to the Japanese butterflies in natural color. Hokuryukan. 361 Jonsson, B. G. 2001. A null model for randomization tests of nestedness in species 362 assemblages. Oecologia 127:309–313. 363 Kaplan, I., and R. F. Denno. 2007. Interspecific interactions in phytophagous insects 364 revisited: a quantitative assessment of competition theory. Ecology Letters 365 10:977–994. 366 367 368 Kneitel, J. M., and J. M. Chase. 2004. Trade-offs in community ecology: Linking spatial scales and species coexistence. Ecology Letters 7:69–80. Koleff, P., K. J. Gaston, and J. J. Lennon. 2003. Measuring beta diversity for 369 presence-absence data. Journal of Animal Ecology 72:367–382. 370 Kozak, K. H., C. H. Graham, and J. J. Wiens. 2008. Integrating GIS-based 371 environmental data into evolutionary biology. Trends in Ecology and Evolution 372 23:141–148. 373 Kubota, Y., T. Hirao, S. Fujii, T. Shiono, and B. Kusumoto. 2014. Beta diversity of 374 woody plants in the Japanese archipelago: the roles of geohistorical and ecological 375 processes. Journal of Biogeography 41:1267–1276. 376 377 Lawton, J. H. and D. R. Strong. 1981. Community patterns and competition in folivorous insects. The American Naturalist 118:317–338. bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 378 379 380 Liu, C., P. M. Berry, T. P. Dawson, and R. G. Pearson. 2005. Selecting thresholds of occurrence in the prediction of species distributions. Ecography 28:385–393. Mayfield, M. M., and J. M. Levine. 2010. Opposing effects of competitive exclusion on 381 the phylogenetic structure of communities. Ecology letters 13:1085–1093. 382 Nakadai, R., and A. Kawakita. 2017. Patterns of temporal and enemy niche use by a 383 community of leaf cone moths (Caloptilia) coexisting on maples (Acer) as revealed 384 by metabarcoding. Molecular Ecology 385 Novotny, V., S. E. Miller, L. Baje, S. Balagawi, Y. Basset, L. Cizek, K. J. Craft, F. Dem, 386 R. A. I. Drew, J. Hulcr, J. Leps, O. T. Lewis, R. Pokon, A. J. A. Stewart, G. Allan 387 Samuelson, and G. D. Weiblen. 2010. Guild-specific patterns of species richness 388 and host specialization in plant-herbivore food webs from a tropical forest. Journal 389 of Animal Ecology 79:1193–1203. 390 Nyman, T., V. Vikberg, D. R. Smith, and J.-L. Boevé. 2010. How common is ecological 391 speciation in plant-feeding insects? A “Higher” Nematinae perspective. BMC 392 Evolutionary Biology 10:266. 393 Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, P. R. Minchin, R. B. O’Hara. 2015. 394 Vegan: community ecology package 395 [https://cran.r-project.org/web/packages/vegan/index.html]. 396 Peterson, A. T., J. Soberon, R. G. Pearson, R. P. Anderson, E. Martinez-Meyer, M. 397 Nakamura, M. B. Araújo 2011. Ecological niches and geographic distributions. 398 Monographs in Population Biology. Princeton University Press. Princeton. New 399 Jersey, USA 400 401 Phillips, S. J., R. P. Anderson, and R. E. Schapire. 2006. Maximum entropy modeling of species geographic distributions. Ecological Modelling 190:231–259. 402 Rabosky, D. L. 2009. Ecological limits and diversification rate: alternative paradigms to 403 explain the variation in species richness among clades and regions. Ecology Letters 404 12: 735– 743. bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 Rathcke, B. J. 1976. Competition and co-existence within a guild of herbivorous insects. Ecology 57:76–87. R Core Team 2015. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/ Ross, H. H. 1957. Principles of natural coexistence indicated by leafhopper populations. Evolution 11:113–129. Saito, M. U., U. Jinbo, M. Yago, O. Kurashima, and M. Ito. 2016. Larval host records of butterflies in Japan. Ecological Research 31:491–491. Schoener, T. W. 1968. The anolis lizards of bimini: Resource partitioning in a complex fauna. Ecology 49:704–726. Schoener, T. W. 1983. Field experiments on interspecific competition. The American Naturalist 122:240–285. Shirôzu, T. 2006. The butterflies of Japan in color. In Japanese. Gakken Holdings, Tokyo, Japan. Stone, L., and A. Roberts. 1990. The checkerboard score and species distributions. Oecologia 85:74–79. Strong, D. R. 1982. Harmonious coexistence of Hispine beetles on Heliconia In experimental and natural communities. Ecology 63:1039–1049. Strong, D. R., J. H. Lawton, and S. R. Southwood. 1984. Insects on plants. Community patterns and mechanisms. Harvard University Press, Cambridge, MA. Tack, A. J. M., O. Ovaskainen, P. J. Harrison, and T. Roslin. 2009. Competition as a structuring force in leaf miner communities. Oikos 118:809–818. Takami, Y., and T. Osawa. 2016. Ecological differentiation and habitat unsuitability maintaining a ground beetle hybrid zone. Ecology and Evolution 6:113–124. Ueckert, D. N., and R. M. Hansen. 1971. Dietary overlap of grasshoppers on sandhill rangeland in northeastern Colorado. Oecologia, 8: 276-295. Waloff, N. 1979. Partitioning of resources by grassland leafhoppers (Auchenorrhyncha, bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 432 433 Homoptera). Ecological Entomology 4:379-385. Warren, D. L., M. Cardillo, D. F. Rosauer, and D. I. Bolnick. 2014. Mistaking 434 geography for biology: inferring processes from species distributions. Trends in 435 Ecology & Evolution 29:572–580. 436 Warren, D. L., R. E. Glor, and M. Turelli. 2008. Environmental niche equivalency 437 versus conservatism: Quantitative approaches to niche evolution. Evolution 438 62:2868–2883. 439 440 441 442 Warren, D. L. 2012. In defense of ‘niche modeling’. Trends in Ecology & Evolution 27:497–500. Webb, C. O., D. D. Ackerly, M. A. McPeek, and M. J. Donoghue. 2002. Phylogenies and community ecology. Annual Review of Ecology and Systematics 33:475–505. 443 444 Supporting information 445 Appendix S1: Table S1. The Japanese butterfly species analyzed in the present study. 446 Appendix S1: Table S2. Summary of the MaxEnt data at the 10-km grid scale. 447 Appendix S1: Table S3. Summary of the MaxEnt data at the 1-km grid scale. 448 Appendix S1: Table S4. Correlations among host use similarity (Host), taxonomic 449 relatedness (Taxon), climate niche similarity (Climate), and total dispersal ability 450 (Dispersal), at two spatial scales. (a) Summary of Mantel test data on pairwise 451 correlations between the explanatory matrices. (b) Summary of partial Mantel test data 452 on standardized Cstd scores between pairs of butterfly species. The “Taxon-Dispersal” 453 data (b) were obtained using the datasets at either grid mesh scale. 454 Appendix S2: Details of the methods used for ecological niche modeling. 455 456 Legend 457 Figure 1. The relationships between Cstd scores and host use similarities (10-km grid 458 scale: Spearman ρ = −0.132, P = 0.0001; 1-km grid scale: Spearman ρ = −0.132, P = bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 459 460 0.0001; Mantel test). bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 461 Figure 1. 462 Cstd score (a) 10km grid scale Host use similarity Cstd score (b) 1km grid scale Host use similarity bioRxiv preprint first posted online Apr. 30, 2017; doi: http://dx.doi.org/10.1101/132530. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license. 463 Table 1. Correlations among host use similarity (Host), taxonomic relatedness (Taxon), 464 climate niche similarity (Climate), and total dispersal ability (Dispersal), at two spatial 465 scales. (a) Summary of Mantel test data on standardized Cstd scores between pairs of 466 butterfly species. (b) Summary of Mantel test data on pairwise correlations between 467 host use similarity and the data of other explanatory matrices. (c) Summary of partial 468 Mantel test data on standardized Cstd scores between pairs of butterfly species. The 469 “Host-Taxon” and “Host-Dispersal” data (b) were analyzed using the datasets of either 470 grid mesh scale. 10-km grid (a) Factor Mantel ρ 1-km grid Mantel ρ P-values P-values Host −0.126 0.0001 −0.126 0.0001 Taxon −0.027 0.0063 −0.037 0.0008 Climate −0.723 0.0001 −0.482 0.0001 Dispersal −0.047 0.0161 −0.072 0.0008 471 10-km grid (b) Mantel ρ Factor 1-km grid Mantel ρ P-values P-values Host-Taxon 0.097 0.0001 0.097 0.0001 Host-Climate 0.209 0.0001 0.240 0.0001 Host-Dispersal 0.067 0.0659 0.067 0.0659 472 (c) 10-km grid Mantel ρ 1-km grid P-values Mantel ρ P-values Factor Covariate Host Taxon −0.124 0.0001 −0.123 0.0001 Host Climate 0.036 0.0020 −0.012 0.3660 Host Dispersal −0.124 0.0001 −0.121 0.0001 473 Spearman's correlation coefficients (ρ values) are shown for all four factors. 474 Bold: P < 0.05; Underlined: 0.05 < P < 0.1 after 9,999 permutations. Cstd score (a) 10km grid scale Host use similarity Cstd score (b) 1km grid scale Host use similarity